Digital video compression system with variable digital filter

Abstract

A method and system for compressing television video signals for digital transmission allows full motion pictures to be transmitted. The analog signal is digitized and stored in a memory A in consecutive pixel value locations. The values of the color components of each pixel value are summed and stored in a memory B. In the second frame, the sum of each pixel value is compared with the sums stored in memory B. If the difference is greater than a filter number, then that particular pixel is transmitted by the control computer. The total number transmitted for each frame is compared to a maximum allowable data rate, and the filter number is adjusted accordingly.

We claim: 
     
       1. A method of compressing television video frames for display, comprising in combination: (a) digitizing a first frame into digital pixel values representing red, green and blue components of each pixel of the first frame;   (b) storing the pixel values in a memory A in consecutive locations;   (c) storing a selected form of the values into a memory B in consecutive locations;   (d) digitizing a next frame into digital pixel values representing red, green, and blue components of each pixel of the next frame;   (e) storing the pixel values of said next frame into memory A, replacing the stored pixel values of the the first frame;   (f) determining the differences between selected forms of the values of said next frame with the selected forms from the first frame previously stored in memory B;   (g) comparing the differences determined in step (f) with a predetermined filter number; and   (h) outputting for display only the pixel values stored in memory A for which the differences determined in step (f) were greater than the filter number.   
     
     
       2. A method of compressing television video frames for display, comprising in combination: (a) digitizing a first frame into digital pixel values representing intensity and color components of the pixel of the first frame;   (b) storing the pixel values in a memory A in consecutive locations;   (c) summing the values of the components of the pixel values of the first frame and loading the sums into a memory B in consecutive locations;   (d) digitizing a next frame into digital pixel values representing intensity and color components of the pixels of the next frame;   (e) storing the pixel values of said next frame into memory A, replacing the stored pixel values of the first frame;   (f) summing the values of the components of the pixel values of said next frame;   (g) determining the differences between the sums determined in step (f) and the sums previously stored in memory B;   (h) comparing the differences determined in step (g) with a predetermined filter number; and   (i) outputting for display only the pixel values stored in memory A for which the differences between the sums determined in step (g) were greater than the filter number.   
     
     
       3. The method according to claim 2 further comprising: totalling the number of pixel values for said next frame which were outputted for display in step (i);   comparing the total number with a prescribed allowable number for display;   decreasing the filter number correspondingly if the total number is less than the prescribed number; and   increasing the filter number correspondingly if the total number is greater than the prescribed number.   
     
     
       4. A method of compressing television video frames for display, comprising in combination: (a) digitizing a frame into digital pixel values representing red, green and blue components of the pixels of the first frame;   (b) storing the pixel values in a memory A in consecutive locations;   (c) summing the values of the components of the pixel values of the frame and loading the sums into a memory B in consecutive locations;   (d) digitizing a next frame into digital pixel values representing red, green and blue components of the pixels of the next frame;   (e) storing the pixel values of said next frame into memory A, replacing the previously stored pixel values fo the preceding frame;   (f) summing the value of the components of the first pixel value of said next frame;   (g) determining the difference between the sum computed in step (f) with the sum stored in memory B for that location;   (h) if the difference determined in step (g) is less than a predetermined filter number, incrementing a counter number;   (i) if the difference determined in step (g) is greater than the filter number, outputting a skip count to a forwarding means, resetting the counter to zero, outputting the pixel value in memory A for that pixel to the forwarding means value, and storing the sum of the pixel value of said next frame in memory B to replace the previous sum;   (j) repeating steps (f)-(i) for each pixel of said next frame; and   (k) forwarding from the forwarding means to a receiver means for display the pixel values and skip counts received by the forwarding means in step (i), with the skip counts locating the forwarded pixel values by informing the receiver means of the number of unchanged pixels between changed pixels.   
     
     
       5. The method according to claim 4 further comprising: totalling the number of pixel values for said next frame which were outputted to the receiver means in step (k);   comparing the total number with a prescribed number allowed for the maximum data rate for the receiver means;   increasing the filter number correspondingly if the total number is greater than the prescribed number; and   decreasing the filter number correspondingly if the total number is less than the prescribed number.   
     
     
       6. The method according to claim 4, further comprising: assigning red, green and blue color values to all bits in the pixel except for the most significant bit; and   indicating by the condition of the most significant bit of a pixel number whether the skip count for the next pixel number being outputted to the receiver means under step (k) is zero; and, if so, outputting to the forwarding means for forwarding only the next pixel value.   
     
     
       7. The method according to claim 4 further comprising: determining if the skip count to be transmitted is less than 128; and, if so, outputting only a one byte skip count to the forwarding means; and   if not, outputting a two byte skip count to the forwarding means and indicating to the receiver means by the condition of the most significant bit of the two byte skip count that it is a two byte skip count.   
     
     
       8. A method of compressing television video frames for transmission over a communications channel, comprising in combination: (a) digitizing a frame into digital pixel values representing red, green and blue components of each pixel of the first frame;   (b) storing the pixel values in a memory A in consecutive locations;   (c) summing the values of the components of the pixel values of the frame and loading the sums into a memory B in consecutive locations;   (d) digitizing a next frame into digital pixel values representing red, green and blue components of each pixel of the next frame;   (e) storing the pixel values of said next frame into memory A, replacing the previously stored pixel values of the preceding frame;   (f) summing the value of the components of the first pixel value of said next frame;   (g) determining the difference between the sum computed in step (f) with the sum stored in memory B for that location;   (h) if the difference determined in step (g) is less than a predetermined filter number, incrementing a counter number;   (i) if the difference determined in step (g) is greater than the filter number, outputting a skip count number to a control computer, resetting the counter to zero, outputting the pixel value in memory A for that pixel to the control computer, and storing the sum of the pixel value of said next frame in memory B to replace to previous sum;   (j) repeating steps (f)-(i) for each pixel of said next frame;   (k) transmitting from the control computer to a receiver the pixel values and skip counts received in step (i), with the skip counts locating the transmitted pixel values by informing the receiver the number of unchanged pixels between changed pixels.   (l) totalling the number of pixel values for said next frame which were outputted to the control computer in step (i);   (m) comparing the total number with a prescribed number allowed for the maximum data transmission rate for the communications channel;   (n) increasing the filter number correspondingly if the total number is greater than the prescribed number; and   (o) decreasing the filter number correspondingly if the total number is less than the prescribed number.   
     
     
       9. A system for compressing television video frames for display, comprising in combination: means for digitizing frames into pixel values representing intensity and color components;   memory A means for storing in consecutive locations the pixel values of a frame;   forwarding means for forwarding pixel values to a receiver means;   output means for outputting selected pixel values in the memory A means to the forwarding means for forwarding;   summing means for summing the values of the components in each pixel value;   memory B means for storing in consecutive locations the sums of the values of the components of the pixel value;   subtraction means for determining the difference between the sums of the pixel values stored in memory B with the sums of the pixel values of a next frame;   comparison means for comparing the difference to a predetermined filter number, and for causing the output means to output the pixel value in memory A to the forwarding means for forwarding if the difference is greater than the filter number; and   counter means for incrementing a new count each time the difference is less than the filter number, and for resetting the count each time the difference is greater than the filter number, to provide a skip count for the forwarding means to forward to the receiver means to identify the locations of the pixel values being transmitted.   
     
     
       10. The system according to claim 9 further comprising: comparison means for comparing the total number of the pixel values forwarded by the forwarding means for a frame to a prescribed number corresponding to the maximum data rate, and for adjusting the filter number correspondingly for each frame.   
     
     
       11. A method for compressing a color television signal on a field to field or frame to frame basis, with said field or frame indicative of a television display consisting of a given number of pixels, each one of said pixels indicative of a given area of said display, comprising the steps of: storing in first separate memory locations the sum of values of the Red, Green and Blue (R,G,B) components for each of said pixels during a first frame,   storing in second separate memory locations a value indicative of the Red, Green and Blue (R,G,B) components for each of said pixels during the next successive frame,   deriving a stored sum for each value stored in said second locations, comparing said sum stored in each of said first locations with said derived sum to provide an output signal for each pixel as stored,   outputting only those pixels whose output signal exceeds a given threshold level selected according to a maximum allowable outputting rate.   
     
     
       12. The method according to claim 11, wherein the step of storing in second memory locations is storing a digital number having a first set of bits indicative of the Red component, a second set of bits indicative of the Green component, and a third set of bits indicative of the Blue component, with at least an additional bit, with said additional bit being a one or zero depending upon whether the next successive pixel is to be updated. 
     
     
       13. The method according to claim 12, further including the step of counting the number of pixels between outputted pixels whose output signal did not exceed said threshold to thereby provide a count between outputted pixels clearly defining the position of each outputted pixel with respect to said display. 
     
     
       14. The method according to claim 13, further including the step of monitoring said counting step to provide a skip count of a given number of digits for pixel counts between outputted pixels which do not exceed a given number, whereby the amount of data necessary to define certain counts is reduced. 
     
     
       15. In a system for compressing color television video frames for display in which system a compressed video signal is provided which signal may be transmitted over a communications link to a remote terminal for displaying said transmitted signal, said system of the type operative to convert a television signal on a frame to frame basis or a field to field basis to a plurality of pixels for each field or frame with said pixels indicative of points or areas on said display with each pixel associated with a digital number indicative of the display contents at said pixel location where said system operates to compare the display content of a first frame with the display content of the next frame to determine a difference in said content as compared to a threshold at each pixel location, to thereby transmit or display update only those pixels associated with said difference, in combination therewith of apparatus for providing said difference in content comprising: first storage means having a plurality of first storage locations each operative to store therein a digital number indicative of the R, G and B content of each pixel of a present frame,   second storage means having a plurality of second storage locations each having stored therein a digital number indicative of the sum of said R, G and B components as added and stored for the frame before said preset frame,   means coupled to said first and second storage means and operative to provide a summed signal from said stored digital number at each pixel location in said first storage means and for comparing said summed signal with said sum stored in the corresponding pixel location, said second storage means to provide an output signal for each pixel in said display, and   means responsive to said output signal for comparing said signal against a variable threshold level to output only those digital numbers as stored in said first storage means at those pixel locations which exceed said threshold level, whereby only the digital values stored in said first storage means indicative of pixel locations exceeding said threshold are outputted.   
     
     
       16. The system according to claim 15, further including means for counting the number of outputted pixels for providing said variable threshold level which varies according to the number of outputted pixels. 
     
     
       17. The system according to claim 16, wherein said digital number stored in said first storage means is a first given number of bits indicative of the Red (R) components, a second given number of bits indicative of the Green (G) component and a third given number of bits indicative of the Blue (B) component. 
     
     
       18. The system according to claim 17, wherein the sum stored in said second storage means represents the digital sum of said first, second and third bits. 
     
     
       19. The system according to claim 17 wherein said digital number as stored in said first storage means further includes at least one additional bit whose value is changed from a zero to a one indicative of the outputting of successive pixels as those pixels which immediately follow one another. 
     
     
       20. A system for compressing a color television signal on a field to field or frame to frame basis, with said field or frame indicative of a television display consisting of a given number of pixels, each one of said pixels indicative of a given area of said display, comprising: storage means having a first plurality of storage locations each separate one associated with a separate pixel in said display and each having stored therein a digital number indicative of the sum of the values of the Red, Green and Blue (R, G, B) components as contained in said television signal during a first frame for each pixel in said first frame, and having a second plurality of storage locations each separate one associated with a separate corresponding pixel whereby each of said second locations corresponds to a separate one of said first locations and having stored in said second locations a digital value indicative of each of the values of said Red, Green and Blue components of a present frame,   logic means coupled to said storage means and operative to compare the sum of said stored components in said first locations for each pixel with the sum of said values stored in said second locations for a corresponding pixel to provide an output signal for each pixel as stored indicative of said comparison exceeding a given threshold value to thereby identify any pixel in said display whose summed value as stored in said first location exceeds the summed value as derived from said second locations and means responsive to said output signal for outputting said digital value as stored in said second location for each of said identified pixels, whereby only those pixel values which exceed said threshold value are outputted and as strictly determined by the content of said R, G and B components as stored.